The 3rd Generation Partnership Project (3GPP) is continuing its work of specifying the Evolved Universal Terrestrial Radio Access Network (E-UTRAN), which consists of the Long Term Evolution (LTE) and System Architecture Evolution (SAE) concepts. In the RAN2 working group, a Study Item on Heterogeneous Network (HetNet) mobility enhancements is ongoing as part of the development of the Release 11 specifications for LTE.
With HetNets, the existing homogeneous network is overlaid with additional lower-power, low-complexity base stations. This approach is expected to mitigate the cost and/or capacity limitations of macro densification and base station upgrades. However, one of the challenges with HetNets is a need to revise the existing mobility procedures for LTE networks to optimize system performance.
A focus area in these efforts to improve mobility procedures for HetNets involves speed estimation, i.e., methods to determine the speed of a particular mobile terminal (a “user equipment” or “UE,” in 3GPP terminology, but often referred to interchangeably as mobile terminal, mobile station, wireless terminal, etc.). According to the current specifications for LTE networks, speed estimation can be determined either in the UE or on the network side. These techniques are referred to herein as “UE-based speed estimation” and “network-based speed estimation,” respectively.
UE-Based Speed Estimation
Mobility State Estimation (MSE) is a concept whereby the UE counts the number of handovers or cell reselections during a particular period of time. MSE can be performed by UEs in both Radio Resource Control (RRC) connected state and in idle mode. In RRC connected state, the MSE can be used to scale certain mobility measurement related parameters. For instance, if so configured by the network, the UE can scale the timeToTrigger parameter of the A3 event, based on the UE mobility state detection. (See the 3GPP document “Radio Resource Control (RRC) Protocol Specification,” 3GPP TS 36.331, v. 10.6.0, July 2012, available at www.3gpp.org.”)
MSE was originally designed to work in homogeneous network deployments. In heterogeneous network deployments, however cells may have widely varying sizes, which mean that UEs moving at the same speed through a given network deployment may get different handover counts, depending on which route they take. This can be seen in FIG. 1, for example, which illustrates a number of macro cells 110, each served by a macro base station 115, overlaid by a number of small cells 120. Assuming that both of the pictured UEs 130 detect both macro cells and small cells as they travel through pictured area, UE A will get a lower handover count (3) than UE B (7) for the pictured routes, even if they move at identical speeds and travel similar distances.
Several proposals have been made regarding how UE-based Mobility State Estimation could be improved to work better in heterogeneous network deployments with varying cell sizes. However, there are a number of issues with UE-based MSE that should be considered. First, common to most proposals is that cell sizes are somehow taken into account when calculating the mobility state. Of course, for this to be done the UE must be informed of cell sizes. This would increase the signaling effort from the network to the UE, and any gain from knowing cell size must be compared with this increased signaling effort. Second, cell size information as standardized now is not accurate and precise enough. This is different for network-based speed estimation, where Operations, Administration, and Maintenance (OAM) configuration can be used to enhance the granularity of this information. Third, the cell shape also affects the MSE. For instance, higher macro cell sectorization will affect the MSE, but the cell is still a macro cell. Again, this would not be apparent to the UE, but could be known at the network level. Finally, Cell Range Expansion will also affect the cell size and thus the MSE.
Network Based Speed Estimation
On the network side, knowledge of the UE speed is an important input to the handover decision, e.g., in order to avoid handover of fast moving UEs to small cells. UE History Information is exchanged between eNBs when a UE is performing handover. More particularly, the Information Element (IE) UE History Information contains information about cells that a UE has been served by, in active state, prior to being handed over to the target cell. The information about the cells includes Global Cell Id, Cell Type (Large, Medium, Small, Very Small) and the time the UE stayed in each cell.
Using this information, the network can form a rough estimate of the UE speed. The estimate can be improved by considering also more detailed information about the cells, such as deployment, position, transmit power, antenna configuration.
To allow further improvements in network side speed estimation, it has been proposed to enhance the Last Visited Cell parameter in the UE History Information IE by adding a more generic speed estimate in kilometers/hour, e.g., in the same format as in a Radio Link Failure (RLF) report. Once a generic format for informing the mobility history between eNBs is available, however, the details of the speed estimation process will be left to network implementation.
Network-based speed estimation only works for UEs that are connected to the network, i.e., in RRC connected state. UEs not connected to the network, i.e., those UEs in idle state, will not report cell changes to the network. As a result, the network has no knowledge of the number of cell borders crossed by idle mode UEs. One challenge for network-based speed estimation is that UEs can frequently move between idle and connected states. This can be the case, for example, with smartphones that are occasionally sending small, isolated packets, with longer duration of inactivity between transmissions. For these users, a problem that arises as a result is that the network cannot gather enough handover statistics for a reliable speed estimate.